Machine vision techniques, such as simultaneous localization and mapping (SLAM), augmented reality (AR), and virtual reality (VR), often rely on the identification of objects within the local environment of a device through the analysis of imagery of the local environment captured by the device. The efficiency of such techniques can be enhanced by identifying a location of the device. However, conventional techniques for locating a device typically do not have sufficient resolution to enhance object identification. Further, the amount of time it takes to identify the device's location limits the utility of the location information.